Although there is strong evidence that control and resolution of human leishmanial infections depends primarily on T cell derived lymphokine mediated activation of parasite infected macrophages, less is known about the nature of the responding cell type(s) which are protective or the Ag(s) that elicit these cells to respond. Studies using whole soluble Ag ("dead Ag") preparations show that patients respond to a wide range of leishmanial Ag. The objective of recent studies has been to characterise the response of T cells from patients with healing or healed cutaneous or mucosal infections to Ag expressed by or derived from actively infected autologous monocytes ("live Ag"). Unfractionated T cells proliferated and produced IFN-gamma in response to both "live" and "dead" Ag. Depletion of CD4+ T cells resulted in the loss of proliferative and IFN-gamma responses to both "live" and "dead" Ag. The effect of CD8 depletion, was variable, and not limited to the cells stimulated by infected monocytes. Expansion of T cells specific for "live" antigens using amastigote infected cells followed by re-stimulation with HPLC fractionated soluble antigens revealed that a diversity of antigens are associated with infected monocytes. There may, however, be quantitative differences in the expression of certain antigens since "live" versus "dead" antigens could distinguish responses between MCL and LCL patients. The immune responses in patients with non-healing, visceral forms of disease have continued to be a focus of study. The critical immunological feature of visceral leishmaniasis is the complete absence of cell mediated immunity to leishmanial antigens. Patients have been shown to have negative intradermal skin tests to leishmania, absent lymphocyte blastogenesis, and decreased IL-2 and IFN-g production in response to parasite antigens. Despite this unresponsiveness, there is evidence that antigen-specific T cells are present in kala-azar patients. Visceral leishmaniasis is accompanied by high titres of antileishmanial antibodies, most of which are likely to be T-cell dependent. Additionally, T-cells are required are likely to be required for successful chemotherapy of visceral disease. Mouse models of cytokine-mediated lymphocyte functional regulation provide a theoretical framework for the present investigation of the antigen-specific anergy of kala-azar. For a descriptive study of cytokine patterns in kala-azar, we sought to maximize assay sensitivity by using semi-quantitative reverse-transcriptase PCR techniques to analyze cytokine mRNA extracted from lesional tissue -- bone marrow. In preliminary data we provide evidence that IL-10 mRNA is present at relatively high levels in active kala-azar, being at or below the limits of detectability after treatment, and absent in uninfected controls.